Improving Gas Sensing Properties of Tin Oxide Nanowires Palladium

0 downloads 0 Views 853KB Size Report
Mar 1, 2013 - sensing response of fabricated sensors. Prepared sensor devices were exposed to liquid petroleum gas (LPG) and vapor of ethanol (C2H5OH) ...
JNS 2 (2013) 469- 476

Improving Gas Sensing Properties of Tin Oxide Nanowires Palladium-Coated Using a Low Cost Technique M. Barzegar*, M. B. Rahmani, H. Haratizadeh Department of Physics, Shahrood University of Technology, Shahrood, 3619995161, Iran Article history: Received 13/1/2013 Accepted 26/2/2013 Published online 1/3/2013 Keywords: Chemical vapor deposition Gas sensor Nanowires Response time Spray pyrolysis

*Corresponding author: E-mail address: [email protected] Phone: 98 935 3915763 Fax: +982733334419

Abstract Thin films of SnO2 nanowires were successfully prepared by using chemical vapor deposition (CVD) process on quartz substrates. Afterwards, a thin layer of palladium (Pd) as a catalyst was coated on top of nanowires. For the deposition of Pd, a simple and low cost technique of spray pyrolysis was employed, which caused an intensive enhancement on the sensing response of fabricated sensors. Prepared sensor devices were exposed to liquid petroleum gas (LPG) and vapor of ethanol (C2H5OH). Results indicate that SnO2 nanowires sensors coated with Pd as a catalyst show decreasing in response time (~40s) to 1000ppm of LPG at a relatively low operating temperature (200oC). SnO2 /Pd nanowire devices show gas sensing response time and recovery time as short as 50s and 10s respectively with a high sensitivity value of ~120 for C2H5OH, that is remarkable in comparison with other reports. 2013 JNS All rights reserved

1. Introduction Applications of chemical gas sensors include environmental monitoring, automotive applications, emission monitoring, and aerospace vehicle health monitoring [1]. Semiconducting metal-oxides have been known for decades to be suitable for gas sensing purposes. There are many reports on applications of these materials as gas sensor devices due to their small dimensions, low cost, and high compatibility with microelectronic processing [2, 3]. Amongst all semiconducting

metal-oxides, Tin oxide (SnO2) is the most widely studied gas sensor material and the most commercially available chemical gas sensors exploit a SnO2 element [4, 5]. Although these oxides themselves are catalytically active, they are rarely used in isolation as their gas sensing characteristics are usually enhanced by using a small amount of noble metal catalyst such as palladium (Pd) and platinum (Pt) [6]. It is widely accepted that the presence of noble metal elements (Pt, Pd, Au, Ag, etc.) on the surface of a metal

470

oxide enhances the interaction of reducing/oxidizing gases with the absorbed oxygen on the surface [3]. Besides improving the selectivity, catalysts also modulate the electron transport properties of the sensing SnO2 layer and improved response characteristics are obtained. The introduction of catalysts influences grain size, the shape of crystallites, bulk and surface stoichiometry, properties of inter-crystalline barriers, and bulk electro-physical properties [7]. Table 1 has summarized maximum response in exposing to LPG in various SnO2 based sensors at different temperatures.

M. Barzegar et al./ JNS 2 (2013) 467-476

high sensitivity with fast response time at a relatively low sensor working temperature.

2. Experimental procedure Thin films of SnO2 NWs were deposited using CVD under a reactive ambient (100sccm of Ar and 15sccm of O2) on top of quartz substrates. Precursor was composed of 0.5gr of SnO2 powder and 0.5gr carbon powder. Alumina boat was used to heat the precursors up to 1050oC with the rate of

7.5oC/min. Afterwards, various concentrations of the solution of palladium chloride (PdCl2) were deposited on the surface of SnO2 nanowire thin films by spray pyrolysis to enhance gas sensing Table 1. Maximum response to LPG in various SnO2- based parameters. Briefly, 1gr PdCl2 powder was solved sensors at different temperatures. in 100cc DI-water and ethanol (1:1 ratio), 0.1mL Sensor type LPG Operating Response HCl was also added to the solution to get a (in ppm) Temperature transparent solution. To find the optimum amount SnO2 Thick film 10,000 350 0.93 [8] of Pd on the surface of SnO2 NWs film, a series of SnO2 Thick film 200 300 0.7 [9] samples with different amounts of 0.02M solution SnO2 Thick film 1000 350 3.68 [10] SnO2 Thick film 800 400 1.38 [11] of PdCl2 (15, 25, 50, and 75cc) were sprayed at SnO2 Thick film 1000 345 0.1 [12] 450oC, as optimised substrate temperature. Finally, SnO2 NWs 500 350 3.5 [5] samples were annealed at 300oC for two hours on SnO2 Hierarchica 500 350 8.1 [5] the flow of 100sccm Ar to get rid of any solvent TGS 2612 Figaro 1000 VH=5V 2.1 [13] traces. Moreover, annealing at 300oC transforms the amorphous SnO2 films into a poly-crystalline In this work, we have synthesized thin films structure [4]. For the fabrication of sensor devices devices comprising Pd-coated SnO2 NWs using and to achieve good contact for the electrical combination of CVD and spray pyrolysis measurements, pairs of Au electrodes were techniques. Effects of Pd as a catalyst on the deposited onto on top of the SnO2/Pd thin film modification of gas sensing properties of SnO2 samples (200nm thickness, 5mm distance. A NWs were studied using a home-made set-up for sample area is 1cm×1cm (Fig. 4(a))), using measuring sensing characteristics. The vacuum thermal evaporation by molybdenum boat. morphologies of the undoped and Pd-doped SnO2 Gas sensing measurements were taken by a testing nanowires were observed by field emission apparatus consisting of a Teflon test chamber with scanning electron microscopy (FESEM; Hitachisa controllable heater, mass flow controllers, and a 4160) and high resolution transmission electron PC-based (PSIP 86D) multimeter. Finally, four microscopy (HR-TEM; JEOL2100F) techniques. kinds of samples with 15, 25, 50 and 75cc sprayed Conductometric gas sensing measurements showed PdCl2 were prepared which indexed as A, B, C and

M. Barzegarr et al./ JNS 2 (2013) 467-476

D, respectivvely and theeir gas sensiing properties were studied toward LPG and vapor of ethanol at a o mperatures (5 50–300 C). different tem

Fig.1. (a) Unndoped SnO2 nanowires n grow wn by CVD onn Quartz substrrate. (b) SnO2 NWs with depposition of 15ccc Pd (sample A). A (c) SnO2 nanowires n withh deposition of o 25cc Pd (sam mple B). (d) Top T view of sample s C withh deposition off 50cc Pd. (e) Sample D withh deposition of o 75cc Pd on SnO S 2 nanowiress.

3. Results and discusssion Fig. 1(a) shhows SEM image i of SnO O2 nanowires deposited with w about 80-100 nm diameter inn average. Fig. 1(c) sho ows nanowiires with Pdd w mean size s of abouut 50nm onn particles with nanowires surface. Fig.. 2(f) showss a HR-TEM M n witth Pd particles image of a single SnO2 nanowire on its surfface (sample B). The hiigh-resolutionn TEM imagee confirms th hat the interpplanar spacingg of Pd particcles attached d to the SnO2 nanowire is about 0.21 nm which iss in agreemeent with otheer reports [14]. The sensing responsse characteriistics of thee Ws sensor strructures weree studied oveer SnO2/Pd NW a temperatture range of o 50 to 3000oC for alll

471

fabricated sensors toward both LPG and vaapor of ethanool. Responsess of all sensorrs show a maxximum value at a particular temperatuure which is called mum temperatture. Fig. 3((b) shows thhe gas optim responnse of un--doped and Pd-doped SnO2 nanow wire thin film ms as a funnction of opeerating tempeerature in thee range from 50 to 300 ◦C C. The testingg gases weree 1000ppm LPG, L in N2 with a total flow f rate of 500 5 sccm. Figg. 4 shows dyynamic responnse of prepaared sensors towards diifferent concentration of C2H5OH vaporr at 100oC. Heere, we r as Sensor use thhe definitionn of sensor response Signal = Rg/RN whhere RN and Rg are the resiistance of thee nanowire exxposed to nitrrogen and thee target gases respectively [3]. A relatiively poor response e for seensor A wass obtained att 100oC to ethanol vapor (Sensor signnal =2.75). Thhe highest response ~120 was also obsserved for sennsor B to 40000ppm ethanool vapor (at 100oC). With sensor perforrmance compaarison of SnnO2-Pd dopeed with pure SnO2 NWs, it can be obbserved that thhere is an opptimum concentration. In thhis case, optiimum concenntration was occurred inn 25cc Pd, which result in s sensor response and remarrkable sensitiivity increasee (Fig 4.). As it depicted in i Figs 4(a) and a 4(c), it can be easilly seen that sensor signall of SnO2-Pd doped NWs has been inccreased by facctors of 24 and a 4 in expposure to 40000 and 8000pppm ethaanol vapoor, respecctively.

4472

M Barzegar et al./ JNS 2 (20113) 467-476 M.

Fig. 2. HR-T TEM images of o (a) - (c) 15ccc Pd depositiion on SnO2 nanowires, n (d) – (f) 25cc Pdd deposition, (gg) – (i) 5 deposition annd (j) – (l) 75ccc Pd depositionn.

473

M. Barzegar et al./ JNS 2 (2013) 467-476

Fig.4. Gas sensing response of fabricated sensors toward ethanol vapor (a) Sensor A at 100oC, (b) Sensor B at 100oC, (c) Sen 100oC and (d) Sensor D at 100oC.

Fig. 5. Gas sensing response of fabricated sensors for LPG sensing response: (a) Undoped SnO2 nanowires at 250oC, (b) Sens 200oC, (c) Sensor B 200oC, (d) Sensor C 200oC and (e) Sensor D 200oC.

4474

M Barzegar et al./ JNS 2 (20113) 467-476 M.

gases can be obtainned from Fig.. 6. Sensor B shows O vapor gas g sensing response r tim me and C2H5OH recoveery time 500s and 10s respectivelyy with sensitiivity value about a 120 which w is remaarkable regardding to previoous reports [9]. The responnse and recoveery time were estimated to t be about 40s 4 and about 90s, respectivvely for sensoor B toward LPG. L Fig. 3. (a) SnO S 2 nanowiress with two golld electrodes onn quartz substrrate. (b) The gass response of unn-doped and Pdddoped SnO2 nanowire thin films f as a functtion of operatingg temperature towards t 1000ppm m LPG.

o Fig 5(a) - (e) indicatees gas sensinng response of L at differrent concentrrations. It cann sensors to LPG be seen thatt sensor B is more m sensitivve to 1000ppm m o LPG at a reelatively low w temperaturee about 200 C than other sensors wiith a responnse value of o d sensors perrformances inn about11. SnnO2-Pd doped the presencce of LPG show s that thhere is not a significant increase in sensor s signal compared too pure SnO2 NWs sensor and only inn the 50cc Pdd s relativee concentratioon, the sensittivity of the sensor to referencee sensor has in ncreased from m 10 to 12. Accordinngly, sensor B was demonnstrated betteer gas responnse toward both targeet gases inn comparison with other fabricated sensors. s It is d thee observed thhat Pd partiicles have decreased working tem mperature of the device in comparisonn with uncoatted SnO2 nan nowires (Fig 5(a) and (c)) which has been b confirmeed by other authors a [6]. Inn the C2H5OH H vapor and LPG L atmosphhere, the betteer sensing respponse of sensor B can bee attributed too the small sizze of Pd partiicles coated on o top of SnO O2 nanowire fillms. Here thee response time (recovery time) is defined as the t time perio od needed forr the device too undergo ressistance chan nging from 10% (90%) too 90% (10%) of the vaalue in equillibrium uponn R andd exposure too the targett gas [3]. Response recovery tim me of fabricaated sensors to t both targeet

6 Response annd recovery behavior of fabbricated Fig. 6. sensorrs to target gasees. (a) Sensor B to 4000ppm ethanol vapor and Sensorr D to 80000ppm at opperating temperrature 100°C and (b) Sennsor A to 80000ppm ethanool vapor and Sensor S C to 40000ppm at 1000°C (c) LPG response of Sennsor A to 10000ppm and Senssor C to 500ppm m at an elevvated temperaature 200°C and a (d) Sensorr B to 1000ppm m LPG and Sennsor D to 10000ppm at operatiing temperaturre 200°C.

s mechhanism, In order to inveestigate the sensing since SnO2 typicallly is an n-tyype wide bannd gap semiconductor, thee electron traansfer resultedd from oxygeen vacancies and the adsoorbed gas mollecules at thee active site on o the surfacce of SnO2 seensors,

475

M. Barzegar et al./ JNS 2 (2013) 467-476

acting as donator or acceptor sites. At working temperature of the device in air, SnO2 sensor adsorbs oxygen (O2-) and moisture (OH-), which traps electrons from the conduction band of SnO2. In the presence of target gases, two typical interactions cause a decrease in resistance of SnO2. First, the target gas chemisorbed at the active site on SnO2 surface. Second, the adsorbed oxidizing agent (O2-) oxidized the target gases, resulting in

to the activation of spill-over of sensing gas molecules by the presence of Pd catalyst particles on to the surface of sensing SnO2 layer. In the C2H5OHvapor and LPG atmosphere, the sensing response of sensor B was much better than those of other sensors. This can be attributed to the smaller size of Pd particles coated on top of SnO2 nanowire films.

electron transferring from target gases to O2-. Most of electrons which have been trapped by O2transfer to the SnO2 surface, which increased electronic conductivity of SnO2 sensors [15]. Upon removal of reducing species or introduction of air or oxygen, the mechanism is reversed and the conductivity returns to its original state. The adsorption of atmospheric oxygen on the sensor surface forms ionic species such as O–2 and O– which acquire electron from the conduction band. The reaction kinetics is as follows [16]: (1) O2 (gas) →O2 (ads) – –2 (2) O2 (ads) + e → O (ads) – – – (3) O (ads) + e → 2O (ads) The oxygen species react with ethanol [17] and LPG [18, 19] through complex series of reactions

4. Conclusion Palladium nanoparticles were deposited on the surface of SnO2 NWs films by spray pyrolysis to enhance gas sensing performances. The undoped and Pd-doped SnO2 hollow nanofibers showed significantly different responses to C2H5OH and LPG according to the sensor temperature and Pd doping concentration. The sensing response characteristics of the SnO2/Pd NWs sensor were studied over a temperature range of 50 to 300oC. The response and recovery time were estimated to be about 40s and about 90s, respectively for LPG. Sensor B with 25cc Pd shows C2H5OH vapor gas sensing response time and recovery time as short as 50s and 10s respectively with high response of about 120 which is remarkable in comparison with

as follows: C2H5OH + 6O– → 3H2O + 2CO2 + 6e– (4) (5) C4H10 + 13O– → 5H2O + 4CO2 + 13e– The above reactions take place only if gases are completely oxidized on the sensor surface. By looking at these reactions, it seems that LPG does not oxidize completely and may be following the reaction scheme given below [20]: (6) C4H10 + 2O– → C4H8 − O + H2O + 2e– Partially oxidized gases may not change the conductivity of sensor element drastically, which might have happened in present study of LPG sensing. The observed enhanced response characteristics for ethanol vapor may be attributed

other reports in this area. The capability of the selective gas sensors was explained in terms of the analyte gases as a function of sensor temperature and Pd doping concentration.

References [1]. R.L. Vander Wal, G.W. Hunter, J.C. Xu, M.J. Kulis, G.M. Berger, T.M. Ticich, Sensors and Actuators B: Chemical 138 (2009) 113 . [2]. K. J. Choi and H. W. Jang. Review and Issues, Sensors 10 (2009) 4083. [3]. J. Huang and Q. Wan. Review and Issues, Sensors 9 (2009) 9903.

476

M. Barzegar et al./ JNS 2 (2013) 467-476

[4]. G. Korotcenkov, I. Boris, A. Cornet, J. Rodriguez, A. Cirera,V. Golovanov, Yu. Lychkovsky, G. Karkotsky, Sensors and Actuators B: Chemical 120 (2007) 657. [5]. Le Viet Thong, Le Thi Ngoc Loan, Nguyen Van Hieu, Sensors and Actuators B: Chemical 150 (2010) 112. [6]. Divya Haridas, Vinay Gupta and K.Sreenivas, Bull. Mater. Sc. 31 (2008) 4.

[11]. Wagh, M.S., Jain, G.H., Patil, D.R., Patil, S.A., Patil, L.A. Sensors and Actuators B: Chemical 122 (2007) 357. [12]. B. Thomas, S. Benoy, K.K. Radha, Sensors and Actuators B: Chemical 133 (2008) 404-413. [13]. TGS 2612 datasheet - for the detection of Methane and LP Gas, www.Figaro.com [14]. Qing Wan, Eric Dattoli, and Wei Lu, Small 4 (2008) 451.

[7]. Divya Haridas, Arijit Chowdhuri, K. Sreenivas, Vinay Gupta, International Journal on Smart Sensing and Intelligent Systems 2 (2009) 12. [8]. Phani, A.R., Monorama, S. And Rao, V.J, Materials Chemistry and Physics 58 (1999) 101. [9]. M.V. Vaishampayan, R.G. Deshmukh, I.S. Mulla, Sensors and Actuators B: Chemical131 (2008) 665. [10]. M. H. Madhusudhana Reddy and A. N. Chandorkar. Thin Solid Films 349 (1999) 260.

[15]. Y. Zhang, et al., Sensors and Actuators B: Chemical 132 (2008) 67. [16]. H. Ji, et al., Journal of Wuhan University of Technology--Materials Science Edition 26 (2011) 661. [17]. P. P. Sahay and R. K. Nath, Sensors and Actuators B: Chemical 133 (2008) 222. [18]. D. N. Suryawanshi, et al., Sensors and Actuators B: Chemical 134 (2008) 579. [19]. B. Adamowicz, et al., Vacuum 82 (2008) 966. [20]. M. Penza, et al., Sensors and Actuators B: Chemical 50 (1998) 52.